Re: [ECOS] Re: CYGNUM_KERNEL_SCHED_TIMESLICE_TICKS undefined

[Gary Thomas <gary at mlbassoc dot com>]

Grant Edwards wrote:
> On 2006-12-03, Bart Veer <bartv@ecoscentric.com> wrote:
>
>   
> >    Grant> I'm trying to build a fairly vanilla "net" template on a
> >    Grant> nios2 target, and I get the following:
> >
> >    Grant> : In function int sched_rr_get_interval(pid_t, timespec*)': /[...]/packages/compat/posix/current/src/sched.cxx:230:
> >    Grant> error: CYGNUM_KERNEL_SCHED_TIMESLICE_TICKS' undeclared (first use this function) /[...]/packages/compat/posix/current/src/sched.cxx:230:
> >    Grant> error: (Each undeclared identifier is reported only once for each function it appears in.)
> >     
>
> [...]
>
>   
> >    Grant> I suppose I should try updating ecosconfig, but it seems
> >    Grant> like the CDL for the POSIX stuff is broken.
> >
> > CYGPKG_POSIX_CLOCKS is a component within the POSIX package,
> > not a top-level package.
> >     
>
> I removed the POSIX package, and the problem went away.
>
>   
> > You caqn disable it by editing the ecos.ecc file or by using
> > the graphical configtool. Only top-level packages can be
> > removed at the command-line.
> >     
>
> So it seems like there's an omission in the POSIX package CDL
> in that it it has an undocumented dependance on timeslicing
> being enabled?
>
>   
> > If you are running under cygwin with an old version of
> > ecosconfig, that segfault may be caused by an underlying
> > problem in cygwin's C++ exception handling. There is a
> > workaround for the problem in the latest ecosconfig sources,
> > or you can grab a prebuilt binary from
> > http://www.ecoscentric.com/devzone/configtool.shtml
> >     
>
> I'm running on Linux, but it is a rather old version of
> ecosconfig.
>
>   
I'm more inclined to think that your source tree is a bit messed up.  
Given that you've had [at least]
two unexplained/unreproduceable problems recently, I think you should 
start by looking at how
you ended up with a tree in such a state.  Perhaps you could try 
comparing a virgin CVS tree with
what you have and see what's different.  The things you can't explain 
may lead you in a direction
to fix thing properly.

n.b. POSIX tests build fine against the public tree and things like the 
missing FLASH tests are
also not an issue.


--
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------


--
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss